Instructions to use FinancialSupport/saiga-7b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use FinancialSupport/saiga-7b with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="FinancialSupport/saiga-7b")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("FinancialSupport/saiga-7b")
model = AutoModelForCausalLM.from_pretrained("FinancialSupport/saiga-7b")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use FinancialSupport/saiga-7b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "FinancialSupport/saiga-7b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "FinancialSupport/saiga-7b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/FinancialSupport/saiga-7b

SGLang

How to use FinancialSupport/saiga-7b with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "FinancialSupport/saiga-7b" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "FinancialSupport/saiga-7b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "FinancialSupport/saiga-7b" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "FinancialSupport/saiga-7b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use FinancialSupport/saiga-7b with Docker Model Runner:
```
docker model run hf.co/FinancialSupport/saiga-7b
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

il saiga è uno strano incrocio di antilopi che vive nelle steppe siberiane. Il nome deriva dal fatto che è un parente di fauno/camoscio e un lontano cugino di cerbero (altri modelli open source ita). E' un progetto portato avanti nei weekend con pochi soldi/tempo a disposizione

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	64.51
AI2 Reasoning Challenge (25-Shot)	63.14
HellaSwag (10-Shot)	83.14
MMLU (5-Shot)	61.66
TruthfulQA (0-shot)	54.99
Winogrande (5-shot)	79.01
GSM8k (5-shot)	45.11

Downloads last month: 1,118

Safetensors

Model size

7B params

Tensor type

BF16

Model tree for FinancialSupport/saiga-7b

Quantizations

3 models

Collection including FinancialSupport/saiga-7b

Dante Alighieri

Collection

a funny little experiment to make LLMs talk like the ancient Italian poet • 4 items • Updated Jan 19, 2024 • 1

Evaluation results

normalized accuracy on AI2 Reasoning Challenge (25-Shot)
test set Open LLM Leaderboard

63.140
normalized accuracy on HellaSwag (10-Shot)
validation set Open LLM Leaderboard

83.140
accuracy on MMLU (5-Shot)
test set Open LLM Leaderboard

61.660
mc2 on TruthfulQA (0-shot)
validation set Open LLM Leaderboard

54.990
accuracy on Winogrande (5-shot)
validation set Open LLM Leaderboard

79.010
accuracy on GSM8k (5-shot)
test set Open LLM Leaderboard

45.110